| FMRS | 59 / 100 · C | 79 / 100 · B |
|---|
| Reliability | 9 / 20 | 12 / 20 |
|---|
| Security and permissions | 13 / 20 | 18 / 20 |
|---|
| Maintenance | 11 / 20 | 16 / 20 |
|---|
| Documentation | 14 / 20 | 18 / 20 |
|---|
| Setup experience | 12 / 20 | 15 / 20 |
|---|
| Best for | - Users of AI coding tools who need full discussion/documentation context rather than just URLs
- Clients that support MCP, such as Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI and Claude Desktop
- Teams deploying a remote MCP service via Docker and Streamable HTTP
- Environments that must egress through a proxy (HTTP/SOCKS5)
| - Engineering teams scaling AI coding agents while controlling spend
- Platform teams that need auditable, rollback-aware agent execution records
- Organizations using multiple agents (Claude, Codex, Gemini) under one governed flow
|
|---|
| Not for | - Users who cannot install a Chromium-based browser on the MCP client machine and mainly need arbitrary-site page_content extraction
- Users who want to expose the remote HTTP port directly to the public internet (it is unauthenticated and unencrypted by default)
- Users unwilling to provide any search provider API key
| - Developers who just want a simple code-completion helper without run governance
- Teams unwilling to accept local JSONL run records with HMAC signing
- Environments without a supported adapter (Claude/Codex/Gemini CLIs, direct provider, or verifier-only)
|
|---|
| Required permissions | - Reads a search provider API key from the environment (one of Serper/SerpBase/Tavily/SearXNG/Sofya/You.com)
- Optional GITHUB_TOKEN (can be read-only and limited to public repositories) for GitHub Issues/Discussions extraction
- Launches a local headless Chromium browser process to fetch arbitrary web pages
- Optional outbound network access via proxy env vars (HTTP_PROXY/HTTPS_PROXY/ALL_PROXY) or KINDLY_CHROME_PROXY
| - Read/write access to the configured repo root (writable scope restrictable via --allow-path/--deny-path)
- Ability to invoke the chosen agent CLI (Claude/Codex/Gemini, etc.) and its underlying model account
- Local disk writes for run records, receipts, and share outputs (share/run-receipt., etc.)
- Execution of user-configured verifier commands (e.g. npm test)
|
|---|
| Risks and side effects | - HTTP transports are unauthenticated and unencrypted by default; exposing the port publicly is risky
- get_content fetches any URL it is given and runs from the server's network location, which can reach internal resources
- Browser reuse shares cookies, local storage, cache and user-agent state across requests
- API keys stored in client config files may leak
- Failed or timed-out fetches may return partial content or an error message
| - Verifier commands are executed; poorly chosen verifier commands carry security risk despite preflight policy checks
- Cost and token figures may be estimates (labeled with provenance) and should not be treated as settled accounting
- Receipt integrity must be verified before use as trustworthy evidence; missing rollback evidence is flagged as EVIDENCE_BOUNDARY
- Policy defaults come from martin.config.yaml and can be overridden by CLI flags; misconfiguration can loosen safety boundaries
|
|---|
| Supported clients | Claude Code, Codex, Gemini CLI, Claude Desktop, Cursor, OpenClaw, Antigravity, VS Code | Claude Code, Codex, Gemini CLI, Cursor, VS Code, GitHub Copilot, Continue |
|---|
| Tools | 2 | 0 |
|---|